llama_model_loader: loaded meta data with 34 key-value pairs and 288 tensors from shieldgemma-2b-IMat-GGUF/shieldgemma-2b.Q8_0.gguf.hardlink.gguf (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = gemma2 llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = Shieldgemma 2b llama_model_loader: - kv 3: general.basename str = shieldgemma llama_model_loader: - kv 4: general.size_label str = 2B llama_model_loader: - kv 5: general.license str = gemma llama_model_loader: - kv 6: general.tags arr[str,1] = ["text-generation"] llama_model_loader: - kv 7: gemma2.context_length u32 = 8192 llama_model_loader: - kv 8: gemma2.embedding_length u32 = 2304 llama_model_loader: - kv 9: gemma2.block_count u32 = 26 llama_model_loader: - kv 10: gemma2.feed_forward_length u32 = 9216 llama_model_loader: - kv 11: gemma2.attention.head_count u32 = 8 llama_model_loader: - kv 12: gemma2.attention.head_count_kv u32 = 4 llama_model_loader: - kv 13: gemma2.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 14: gemma2.attention.key_length u32 = 256 llama_model_loader: - kv 15: gemma2.attention.value_length u32 = 256 llama_model_loader: - kv 16: general.file_type u32 = 7 llama_model_loader: - kv 17: gemma2.attn_logit_softcapping f32 = 50.000000 llama_model_loader: - kv 18: gemma2.final_logit_softcapping f32 = 30.000000 llama_model_loader: - kv 19: gemma2.attention.sliding_window u32 = 4096 llama_model_loader: - kv 20: tokenizer.ggml.model str = llama llama_model_loader: - kv 21: tokenizer.ggml.pre str = default llama_model_loader: - kv 22: tokenizer.ggml.tokens arr[str,256000] = ["", "", "", "", ... llama_model_loader: - kv 23: tokenizer.ggml.scores arr[f32,256000] = [-1000.000000, -1000.000000, -1000.00... llama_model_loader: - kv 24: tokenizer.ggml.token_type arr[i32,256000] = [3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, ... llama_model_loader: - kv 25: tokenizer.ggml.bos_token_id u32 = 2 llama_model_loader: - kv 26: tokenizer.ggml.eos_token_id u32 = 1 llama_model_loader: - kv 27: tokenizer.ggml.unknown_token_id u32 = 3 llama_model_loader: - kv 28: tokenizer.ggml.padding_token_id u32 = 0 llama_model_loader: - kv 29: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 30: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 31: tokenizer.chat_template str = {{- bos_token }}\n{%- if messages[-1].... llama_model_loader: - kv 32: tokenizer.ggml.add_space_prefix bool = false llama_model_loader: - kv 33: general.quantization_version u32 = 2 llama_model_loader: - type f32: 105 tensors llama_model_loader: - type q8_0: 183 tensors llm_load_vocab: special tokens cache size = 249 llm_load_vocab: token to piece cache size = 1.6014 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = gemma2 llm_load_print_meta: vocab type = SPM llm_load_print_meta: n_vocab = 256000 llm_load_print_meta: n_merges = 0 llm_load_print_meta: vocab_only = 0 llm_load_print_meta: n_ctx_train = 8192 llm_load_print_meta: n_embd = 2304 llm_load_print_meta: n_layer = 26 llm_load_print_meta: n_head = 8 llm_load_print_meta: n_head_kv = 4 llm_load_print_meta: n_rot = 256 llm_load_print_meta: n_swa = 4096 llm_load_print_meta: n_embd_head_k = 256 llm_load_print_meta: n_embd_head_v = 256 llm_load_print_meta: n_gqa = 2 llm_load_print_meta: n_embd_k_gqa = 1024 llm_load_print_meta: n_embd_v_gqa = 1024 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-06 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 9216 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 10000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_ctx_orig_yarn = 8192 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: model type = 2B llm_load_print_meta: model ftype = Q8_0 llm_load_print_meta: model params = 2.61 B llm_load_print_meta: model size = 2.59 GiB (8.50 BPW) llm_load_print_meta: general.name = Shieldgemma 2b llm_load_print_meta: BOS token = 2 '' llm_load_print_meta: EOS token = 1 '' llm_load_print_meta: UNK token = 3 '' llm_load_print_meta: PAD token = 0 '' llm_load_print_meta: LF token = 227 '<0x0A>' llm_load_print_meta: EOT token = 107 '' llm_load_print_meta: max token length = 48 ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA GeForce RTX 4090, compute capability 8.9, VMM: yes llm_load_tensors: ggml ctx size = 0.26 MiB llm_load_tensors: offloading 26 repeating layers to GPU llm_load_tensors: offloading non-repeating layers to GPU llm_load_tensors: offloaded 27/27 layers to GPU llm_load_tensors: CPU buffer size = 597.66 MiB llm_load_tensors: CUDA0 buffer size = 2649.78 MiB .................................................................. llama_new_context_with_model: n_ctx = 512 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 10000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CUDA0 KV buffer size = 52.00 MiB llama_new_context_with_model: KV self size = 52.00 MiB, K (f16): 26.00 MiB, V (f16): 26.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.98 MiB llama_new_context_with_model: CUDA0 compute buffer size = 504.50 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 6.51 MiB llama_new_context_with_model: graph nodes = 1050 llama_new_context_with_model: graph splits = 2 system_info: n_threads = 25 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | compute_imatrix: tokenizing the input .. compute_imatrix: tokenization took 124.368 ms compute_imatrix: computing over 128 chunks with batch_size 512 compute_imatrix: 0.53 seconds per pass - ETA 1.12 minutes [1]9.0357,[2]6.0710,[3]5.7765,[4]7.3202,[5]7.5810,[6]6.5524,[7]7.3620,[8]7.7717,[9]7.9980, save_imatrix: stored collected data after 10 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [10]7.0250,[11]7.1243,[12]7.7515,[13]8.3860,[14]8.5456,[15]9.1350,[16]9.4929,[17]9.6020,[18]10.0479,[19]9.6072, save_imatrix: stored collected data after 20 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [20]9.6732,[21]10.0235,[22]9.9405,[23]10.1465,[24]10.4632,[25]10.6809,[26]10.4293,[27]10.8468,[28]11.1851,[29]11.2025, save_imatrix: stored collected data after 30 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [30]11.2146,[31]10.5221,[32]10.1902,[33]9.9792,[34]9.7741,[35]9.5936,[36]9.7059,[37]9.7916,[38]9.8392,[39]10.0398, save_imatrix: stored collected data after 40 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [40]10.2194,[41]10.4468,[42]10.9108,[43]11.3623,[44]11.7863,[45]12.0862,[46]11.8589,[47]11.8716,[48]12.1056,[49]12.2897, save_imatrix: stored collected data after 50 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [50]11.9840,[51]11.9845,[52]12.0356,[53]12.1869,[54]12.4523,[55]12.6451,[56]12.6647,[57]12.6160,[58]12.6291,[59]12.4129, save_imatrix: stored collected data after 60 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [60]12.2665,[61]12.0820,[62]12.0084,[63]12.1248,[64]12.1068,[65]12.0451,[66]12.0694,[67]12.0037,[68]11.9044,[69]11.9355, save_imatrix: stored collected data after 70 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [70]11.9182,[71]11.8898,[72]11.8851,[73]11.8183,[74]11.7600,[75]11.6969,[76]11.6840,[77]11.7161,[78]11.6904,[79]11.6315, save_imatrix: stored collected data after 80 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [80]11.6993,[81]11.7500,[82]11.7088,[83]11.7027,[84]11.7604,[85]11.5913,[86]11.5599,[87]11.4877,[88]11.4697,[89]11.4879, save_imatrix: stored collected data after 90 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [90]11.4918,[91]11.3864,[92]11.2760,[93]11.1488,[94]11.0230,[95]10.9294,[96]10.8244,[97]10.7222,[98]10.6331,[99]10.6455, save_imatrix: stored collected data after 100 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [100]10.6567,[101]10.7950,[102]10.8930,[103]10.9870,[104]11.2051,[105]11.3753,[106]11.3982,[107]11.4249,[108]11.4346,[109]11.4167, save_imatrix: stored collected data after 110 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [110]11.3967,[111]11.3270,[112]11.2420,[113]11.2789,[114]11.2914,[115]11.2924,[116]11.2687,[117]11.3088,[118]11.3315,[119]11.3342, save_imatrix: stored collected data after 120 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat [120]11.3327,[121]11.3302,[122]11.2633,[123]11.3580,[124]11.4479,[125]11.5279,[126]11.6469,[127]11.7485,[128]11.8454, save_imatrix: stored collected data after 128 chunks in shieldgemma-2b-IMat-GGUF/imatrix.dat llama_print_timings: load time = 1270.86 ms llama_print_timings: sample time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) llama_print_timings: prompt eval time = 42475.45 ms / 65536 tokens ( 0.65 ms per token, 1542.91 tokens per second) llama_print_timings: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) llama_print_timings: total time = 44660.10 ms / 65537 tokens Final estimate: PPL = 11.8454 +/- 0.20096